Indeed, deaths get reported that's why the death count is the most reliable. Deaths are a lagging indicator of total infection cases because it takes time for

Indeed, deaths get reported that's why the death count is the most reliable. Deaths are a lagging indicator of total infection cases because it takes time for an infection to cause death. Thus if you want to derive infections from deaths you can, but not a current estimate. The US has vastly under tested, so if you want to derive an estimate of current totals you can only go based on lagging indicators. Likewise you need to then forward project an exponential. None of this suggest accuracy, but it will be closer to the ground truth than testing derived numbers are now.